In some network environments, data is periodically transferred from one data processor to another data processor. Two examples of instances involving data transfer are data replication and data backup.
Data replication is the process of sharing changes between multiple copies of a single file or data set on multiple data processing systems. For example, two data processing systems can communicate through a network. Both systems on the network can have a copy of the same data. As time progresses, each data processing system updates its copy. After a period of time, the data may not be the same between data processing systems. Replication can be performed to synchronize the data, and any updates to the data can be merged into one common data set.
Accordingly, data replication can allow users on various networks in a variety of locations to share the same information and files. Replication can be performed with replication software applications in a variety of processor relationships, such as client-server, server-server, and peer-peer replication.
Data replication is typically scheduled to occur at specified times. For example, replication can be scheduled to occur at a predetermined frequency, such as every two hours Monday through Friday and every four hours on weekends. However, the scheduling of data replication can disrupt system operations by using system resources. Occasionally, the scheduled data replication can occur at the same time that a data processing system is experiencing a heavy load. During such poorly timed replications, applications that may be running on the data processing system can be slowed. Response time for connecting users to the system may also be increased. In some instances, users may be unable to connect to the data processing system. The performance of applications sharing the same processor resources may also be adversely affected. Occasionally, the diversion of additional resources needed to perform data replication can result in an application failure as a result of low system resources.
Data transfers to back up data typically involve storing a copy of data separately from an original copy. For example, a document on a client data processing system can be saved periodically on a server processor. Many data backup systems are two tiered environments having a client machine with data that is backed up on a server machine. The client is scheduled to send the data to the server at prescheduled times, and the server manages the backed up data.
Data transfers to back up data typically occur at regularly scheduled intervals. If the data backup process is scheduled during a period of high processor use, applications, connection times, and other system performance measures may be adversely effected.